Identifying predictors of continuance intention on social media-based physical activity using the analytic hierarchy process method

The world population is being threatened by immense rates of physical inactivity in the form of health issues such as cardio diseases, obesity, and etc. Physical activity (PA) such as walking, cycling, cleaning houses and washing cars can improve healthy mind and body. Social media (SM) is a growing social-networking tool connecting people from different states across the globe. Plus, SM has great potential to increase PA level from meta-analyses they exhibited the possibility of changing behavior and many sedentary lifestyles, for example watching TV, playing games, and working with computers can be reduced with the help of SM. Whereby, maintaining PA behavior can turn into a sophisticated topic as people would need constant motivations. The benefits of successfully predicting and understanding continuance intention (CI) will give us a clear picture of what the significant factors are. The objective of this study is to build a model that predicts and understands how Facebook (FB) users' CI for PA developed by analyzing the prior works and the existing theoretical theories. Modeling methods used in this work are Wordcloud, the analytic hierarchy process's calculation (AHP), the review of cognitive theories, and the synthesis of the base models. The CI model is comprised of the same theory of planned behavior's (TPB) constructs like attitude, intention, perceived behavioral control (PBC) with perceived value (PV), as the additional construct and technology acceptance model's (TAM) constructs like perceived usefulness (PU) and ease of use (PEOU). Nevertheless, the CI model also needs the extension of social network factors, therefore, social network structure and characteristics of network ties are included to measure the impact of social influence as a replacement of subjective norm. The findings are fundamental to understanding of the mechanisms driving CI on SM-based PA and applicable to domains like epidemiology, public health and so on.